BROWSE ALL ARTICLES BY TOPIC
Determination of Soluble Vitamins in Beverages
HPLC is a method with simple sample preparation
by Deanna C. Hurum, Brian M. De Borba, and Jeffrey S. Rohrer
Functional beverages are vitamin-enhanced waters that are popular with consumers because of convenience, perceived health benefits, and improved flavor over tap water. These beverages, enriched with vitamin C, B-complex vitamins, and vitamins A and E, are promoted as offering the benefits of increased energy from B vitamins and antioxidant value from vitamins A, C, and E. Sales of these beverages are expected to increase to 4.4 billion liters per year by 2011.1
Because the U.S. Food and Drug Administration (FDA) regulates how the nutritional content is listed on these beverages, vitamin assay methods are needed to support product labeling.2 Determination of vitamins in foods is inherently difficult, and deviation of the determined amounts from the labeled values has been observed.3 Analysis of these waters is challenging due to the presence of both water- and fat-soluble vitamins. Proprietary formulations of vitamins that remain soluble and shelf-stable are used to enrich these beverages. Additionally, gums, preservatives, and other additives are used to emulsify and stabilize the drink.
Traditional determinations require that multiple methods be used to quantify the various vitamins added to the beverage. Water-soluble vitamins are often determined with reversed phase-high performance liquid chromatography (HPLC) using an aqueous mobile phase, while the fat-soluble vitamins use organic solvent mobile phases in both reversed- and normal-phase HPLC methods.4 Simultaneous determination of both types of vitamins poses a challenge due to the difference in solubility limits of the two classes of vitamins and the many different biologically equivalent compounds that can be added but are listed as a single vitamin. For example, niacin is available as nicotinic acid and nicotinamide, both of which are biologically active and referred to as niacin in product labeling.
The simultaneous determination of a wide range of vitamins increases the complexity of an analytical method. Their structures range from small unconjugated organic acids that are minimally ultraviolet (UV) active, such as pantothenic acid (vitamin B5), to large complexes that absorb at different wavelengths, such as cyanocobalamine (vitamin B12). Due to the chemical diversity of vitamins, multiple detection wavelengths are needed to optimize the method sensitivity.
Described below is a single method with simple sample preparation that can determine many common vitamins added to functional waters.5 By using a silica-based polar-embedded column compatible with 100% aqueous mobile phases over a wide pH range (1.5-10), a single gradient method can be used to quantify both water- and fat-soluble vitamins with stable retention times, excellent peak shapes, and high efficiencies for basic and acidic compounds.
Column: Acclaim PolarAdvantage II 3 µm, 2.1 x 150 mm
Gradient: Mobile phase A: 0.015% formic acid in deionized (DI) water Mobile phase B: 17/83 methanol/acetonitrile 100% A for three min, 0–45% B in five min, 45–100% B in 0.1 min, 100% B for 16.9, five min of equilibration at 100% A prior to injection
Flow Rate: 0.21 mL/min
Inj. Volume: 5 µL
Detection: Photodiode array detector; 210, 280, and 350 nm
Three fruit-flavored vitamin-enhanced waters were analyzed for vitamin content. Samples were diluted 1:1 with 0.015% formic acid prior to analysis.
Water-Soluble Vitamin Standards: Vitamin standards of pyridoxine HCl, nicotinic acid, nicotinamide, D-pantothenic acid, and cyanocobalamine were prepared by weighing 10 mg to 20 mg of the vitamin powder and adding DI water to a total of 10 g to 20 g to form a stock solution of one mg/mL for each individual vitamin. Water-soluble vitamin stock solutions were stored at -20°C when not in use. Working standards containing vitamins in 0.015% formic acid (mobile phase A) were prepared from these stocks on the day of use.
Fat-Soluble Vitamin Standards: The fat-soluble vitamins were prepared in acetonitrile to yield a one mg/mL solution of DL-α-tocopherol acetate and a 0.2 mg/mL solution of retinol palmitate. Retinol palmitate requires several minutes of vortexing to dissolve. These solutions were stored in the dark at 4°C . Calibration standards of retinol palmitate and DL-α-tocopherol acetate were not prepared in a matrix of 0.015% formic acid. Instead, due to solubility limitations, working standards of these vitamins were prepared in mobile phase B from stocks prepared in acetonitrile, and a separate standard curve was prepared for the fat-soluble vitamins.
Nine Vitamins Separated
Figure 1 (see p. 34) shows the separation of nine vitamins with the proposed method. Sodium citrate and citric acid, common ingredients that enhance tartness, were added to the mixture to demonstrate the separation between nicotinamide and the citrate peak. As shown in Figure 1, citrate and the vitamins are well separated from each other.
Vitamins are a structurally diverse group of compounds with different absorbance maxima. For example, retinol palmitate has a strong absorption at 350 nm, while D-pantothenic acid absorbs in the UV between 200 nm and 225 nm. Given the wide UV absorbance range for the vitamins commonly added to enhanced waters, wavelengths of 210 nm, 280 nm, and 350 nm were used for detection. In Figure 1, the chromatogram collected at 210 nm captures each of the vitamins except for vitamin A.
The linearity, limit of detection, limit of quantitation, and precision data for this gradient method were determined for nine vitamins: pyridoxine HCl (B6), ascorbic acid (vitamin C), nicotinic acid (B3), nicotinamide (B3), pantothenic acid (B5), cyanocobalamine (B12), folic acid, tocopherol acetate (vitamin E), and retinol palmitate (vitamin A).
Correlation coefficients ranged from 0.9985 to 0.9997. Retention time precisions (relative standard deviations or RSDs) ranged from 0.07% to 0.23% with peak area precisions (RSDs) ranging from 0.28% to 0.97%, with the exception of ascorbic acid, which had a peak area precision of 3.47%. Ascorbic acid exists in solution in equilibrium with the oxidation product dehydroascorbic acid (DHAA). While both compounds are biologically active as vitamin C, DHAA does not strongly absorb in the UV and is therefore difficult to quantify. The oxidation reaction is reversible and is minimized at low pH or by adding a reducing agent, such as dithiotheitol, to prevent oxidation of ascorbic acid by dissolved oxygen.6
Functional Beverage Samples
Three brands of vitamin-enhanced water were analyzed over three days to evaluate the precision of the method. Representative data are presented in Figure 2 (see above) with a summary in Table 1 (see left). Intra-day retention time precision ranged between 0.04% and 0.23%, which is equivalent to the precision of the standards. Inter-day peak area precision ranged from 0.37% to 9.5%. The increased imprecision observed in the folic acid results is due not only to the presence of a closely eluting peak in the brand B sample but also to the low concentration of folic acid in the sample. Even though this is a challenging matrix and there is a low concentration of folic acid, it is still easily quantified.
Recoveries for the water-soluble vitamins ranged from 93% to 119%. As with precision, the extremes in recoveries were for ascorbic acid and folic acid. Despite the challenges in quantifying these two vitamins, the recoveries were good, showing that the method is accurate.
Direct spiking of the acetonitrile stocks of vitamins A and E into vitamin-enhanced waters led to the formation of an unstable suspension; therefore, a control sample was used to indirectly measure recovery. Recoveries for fat-soluble vitamins were determined by comparison of spiked samples to a control sample of retinol palmitate and DL-α-tocopherol acetate in 0.015% formic acid, leading to recovery values ranging from 101% to 110%.
As an additional check, the determined values (DV) of vitamins A and E were compared to the label claim. Brand A claimed to provide 10% of the DV of vitamin E per serving, and brand B claimed 20% of the DV of vitamin E. In this study, we determined 15% and 25% of vitamin E in Brands A and B, respectively. Brand C claimed 10% each of vitamins A and E. The determined amounts were 10% each, based on FDA guidelines for nutritional labels. In the absence of a formulated aqueous-soluble mixture of the lipophilic vitamins, these results agree well with the label claim and show suitability of the method to determine these vitamins in beverages.
This work demonstrates that a single HPLC method using minimal sample preparation could be used to determine water- and fat-soluble vitamins in an enhanced water sample with high reproducibility. This was possible because of a polar-embedded stationary phase that allowed a gradient of 100% to 0% aqueous mobile phase without loss of column efficiency. Additionally, this separation method is compatible with mass spectrometry, allowing for a complementary detection method.
This single-method determination of both water- and fat-soluble vitamins has good linearity, precision, recovery, and limits of quantitation. ■
- Zegler J. State of the industry-beverage blockbusters 2007. Beverage Industry website. February 8, 2008. Available at: www.bevindustry.com/Archives_Davinci?article=2140. Accessed 5/14/2010.
- U.S. Food and Drug Administration. Title 21 Code of Federal Regulations (21 CFR Part 101). Food labeling. Available at: http://ecfr.gpoaccess.gov/cgi/t/text/text-idx?c=ecfr&tpl=/ecfrbrowse/Title21/21cfr101_main_02.tpl. Accessed 5/14/2010.
- Sharpless KE, Margolis S, Thomas JB. Determination of vitamins in food-matrix Standard Reference Materials. J Chromatogr A. 2000;881(1-2):171-181.
- Jedlicka A, Klimes J. Determination of water- and fat-soluble vitamins in different matrices using high-performance liquid chromatography. Chem Pap. 2005;59(3):202-222.
- Dionex Corporation. Determination of water- and fat-soluble vitamins in functional waters by HPLC with UV-PDA detection. Application Note 216. Available at: http://www.dionex.com/en-us/webdocs/71938-AN216-HPLC-Vitamins-FunctionalWaters-30April09-LPN2145.pdf. Accessed May 14, 2010.
- Margolis SA, Duewer DL. Measurement of ascorbic acid in human plasma and serum: stability, intralaboratory repeatability, and interlaboratory reproducibility. Clin Chem. 1996;42:1257-1262.